Using Ecological Momentary Assessment to Examine Perceptions of Safety, Aesthetics, and Physical Activity in Adults

Transcription

1 Using Ecological Momentary Assessment to Examine Perceptions of Safety, Aesthetics, and Physical Activity in Adults Genevieve F. Dunton, PhD, MPH Christina Younan, B.S. Candidate Yue Liao, MPH Keito Kawabata, B.A. Donna Spruijt-Metz, Ph.D Stephen Intille, Ph.D Mary Ann Pentz, Ph.D

2 Perceived Environmental Features and Physical Activity Vegetation/Greenness (Tilt et al., 2007; Sugiyama et al., 2008) Aesthetics (Inoue et al., 2011; Saelens et al., 2011) Safety (Zoellner et al., 2012; Tucker-Seeley et al., 2009) Garbage/Physical Disorder (Corseuil et al., 2011) Traffic Volume ([NS] Hoehner et al., 2005; [NS] McGinn et al 2007)

3 Methodological Limitations Assessments not conducted in the target settings Multiple micro settings condensed into one rating May assess settings that are never or not regularly encountered Recall errors or reporting biases Perceptions of settings may differ when perceiver is physically active versus inactive Use of settings is not measured

4 Ecological Momentary Assessment (EMA) Real-time responses in naturalistic settings Can simultaneously measure: 1) Specific location (e.g., backyard, trail, sidewalk) 2) Perceived characteristics (e.g., safety, traffic) 3) Behavior (e.g., sports/exercise, eating) Ecologically valid Less recall bias

5 Study Objectives 1) To determine whether perceived vegetation, traffic, safety, and litter/garbage are associated with physical activity levels in those settings. 2) To determine whether these associations between environmental perceptions and physical activity differ for males versus females.

6 Participants N 59 Age M = 39.9 (SD = 9.8) (range: years) Sex Ethnicity 74% Female 33% Hispanic 29% White/Caucasian 24% Asian 7% African-Am. 7% Other Income 26% $40,000 26% $100,000 Weight Status 56% Overweight/Obese

7 EMA Equipment Mobile phone (HTC Shadow, T-Mobile)

8 EMA Prompting Schedule 32 prompts occurred across 4 days (Sat-Tue) Ecological Momentary Assessment Prompting Schedule Day 6: am 10am- 12-2pm 2-4pm 4-6pm 6-8pm 8-10pm 6:45am 12pm Saturday X X X X X X X X Sunday X X X X X X X X Monday X X X X X X X X Tuesday X X X X X X X X Note: Question sequences are prompted at a random time within each interval.

9 EMA Protocol Up to 2 reminder prompts after 3 min for missed and unfinished entries. Phone required nightly charging. One contact call and one reminder text from project staff on Mon or Tue. Study hotline available for technical problems. Participants paid up to $50 ($18 for returning phone and $1 x 32 for each complete survey)

10 EMA Items

11 Accelerometer Actigraph GT2M. Time-stamped and linked with EMA data. Outcome variable: Total steps in the 15-min. of EMA prompt

12 Predicted marginal means SUDAAN 10.0 Data Analyses Generalized Estimating Equations (GEE) regressions adjusted the SE s for the clustering of EMA observations within each person Unit of analysis was the EMA observation All models controlled for age, gender, and income

13 EMA Compliance and Missing Accelerometer Data On average, participants responded to 80% (range 25% - 100%) of EMA prompts. Percentage of missing data unrelated to age, sex, income, ethnicity, and BMI. Number of steps ( 15-min.) did not differ for unanswered vs. answered EMA prompts. On average, 21% of answered EMA prompts could not be matched due to accelerometer non-wear (more likely at home [indoors]).

14 Steps (+/- 15 min.) Activity Levels by Physical Context Home (Indoors) Work (Indoors) Home (Outdoors) Outdoors (not at home) Car Other

15 Results of GEE Regression Models Predicting Steps ( 15 min) Main Effect β (SE) Interaction By Gender β (SE) Vegetation 77.77** (24.07) ** (38.13) Safety (69.70) (54.51) Traffic (28.44) (139.13) Litter/Garbage ^ (36.81) (68.68) Note. **p <.01, ^p <.10. All models control for age and annual household income

16 Steps (+/- 15 Min.) Perceived Vegetation by Gender Interaction for Activity Levels Males Females Perceived Vegetation

17 Steps (+/- 15 Min.) Association Between Perceived Litter/Garbage and Activity Levels Perceived Litter/Garbage

18 Conclusions Physical activity levels are higher in locations with greater perceived vegetation - Rules out neighborhood third variable problem - Females may have greater preference to be active in greener areas Perceived litter/garbage may deter physical activity

19 Limitations Perceptions not assessed during every physical activity episode Missing data Short monitoring period (4 days) 1-item measures Statistical power

20 Acknowledgments Programming: Jennifer Beaudin (MIT) Project Manager: Keito Kawabata American Cancer Society Mentored Research Scholar Grant MRSGT CPPB (Dunton, PI). National Cancer Institute #R01-CA (Pentz, PI). ALR Accelerometer Loan Program

21 Thank You